#pentagram@irc.freenode.net logs for 16 Jun 2003 (GMT)

Archive Today Yesterday Tomorrow
Pentagram homepage


[01:50:39] --> Cash|Away has joined #pentagram
[02:02:10] <-- Kirben has left IRC (Read error: 54 (Connection reset by peer))
[02:10:52] --> Kirben has joined #pentagram
[02:10:52] --- ChanServ gives channel operator status to Kirben
[02:15:27] <-- Cash|Away has left IRC ()
[03:49:53] <-- Kirben has left IRC (kornbluth.freenode.net irc.freenode.net)
[03:50:19] --> Kirben has joined #pentagram
[04:50:41] <-- Kirben has left IRC (Read error: 104 (Connection reset by peer))
[04:57:58] --> Kirben has joined #pentagram
[04:57:58] --- ChanServ gives channel operator status to Kirben
[05:08:11] <-- servusAway has left IRC (Read error: 110 (Connection timed out))
[05:13:36] <-- Kirben has left IRC (Read error: 104 (Connection reset by peer))
[05:20:03] --> Kirben has joined #pentagram
[05:20:07] --- ChanServ gives channel operator status to Kirben
[06:19:33] --> Cashman has joined #pentagram
[06:20:12] * Cashman updates so he can check out the new stuff
[06:20:30] * Cashman thinks about helping submitt bugs
[06:29:26] <Cashman> compiling.... hmmm
[06:38:23] <Cashman> hmm dev ide's got problems with nested header files again *hick*
[06:38:25] <Cashman> fixing
[06:48:10] <Cashman> gumps/GameMapGump.cpp:152: no matching function for call to `Item::getZ()'
[06:48:23] <Cashman> where is this function located? in which header
[06:58:46] <-- Cashman has left IRC ()
[07:07:22] --> wjp has joined #pentagram
[07:07:22] --- ChanServ gives channel operator status to wjp
[07:07:28] <wjp> hi
[07:46:52] --- wjp is now known as wjp|work
[08:01:22] --> Cashman has joined #pentagram
[08:03:18] <Cashman> 107 C:\Nick\Pentagram\pentagram\world\Item.cpp
[08:03:18] <Cashman> no `sint32 Item::getZ() const' member function declared in
[08:03:27] <DarkeSleep> Cashman: Check out the cvs again if you haven't located the problem. It's possible you got a download halfway through an upload.
[08:03:37] <Cashman> ok
[08:03:54] <Cashman> I see yeah I know its happened before!
[08:04:28] <DarkeSleep> wjp also broke cvs lastnight too, so it may be his fault. *grin*
[08:04:49] <Cashman> yeah it might be that
[08:05:24] <Cashman> I did read the logs I see you have some collision detection, so do the roofs disappear when you walk into houses now?
[08:05:47] <DarkeSleep> For reatively unique definitions of 'walk', yes. *grin*
[08:06:10] <Cashman> yeah hehe "can't avatar in statis"
[08:07:05] <wjp|work> this is most likely the fault of SF's daily anon-cvs updates
[08:07:05] <Cashman> ok so are there specific shapes you looked for? maybe box's
[08:07:13] <Cashman> ok
[08:07:22] <wjp|work> check SF's "site status" page for details
[08:07:34] <wjp|work> roof popping: shapes have a 'roof' flag that I check for
[08:07:50] <wjp|work> (a lot better than U7 which just had a hardcoded list of shape numbers :-) )
[08:08:11] <wjp|work> CVS currently compiles fine here
[08:08:14] <Cashman> ok so there is a separate flags file like the usecode
[08:08:16] <wjp|work> (but that's developer CVS, not anon cvs)
[08:08:21] <Cashman> ok
[08:08:31] <wjp|work> typeflag.dat
[08:08:40] <Cashman> true
[08:08:42] <wjp|work> 8 bytes per shape
[08:08:58] <wjp|work> contains dimensions of shapes, weight, volume, flags, things like that
[08:09:26] <Cashman> ok nice! thats gotta be a help
[08:09:55] <Cashman> you just started work?
[08:10:02] <Cashman> well a little while ago?
[08:10:06] <wjp|work> about an hour ago
[08:10:08] <wjp|work> (9am)
[08:12:17] <Cashman> yeah I was wondering about the flags before - much better than hardcoded for sure :-)
[08:15:40] <Cashman> no `sint32 Item::getZ() const' member function declared in
[08:15:40] <Cashman> class `Item'
[08:15:51] <Cashman> wjp when you get to it, any ideas what happened there
[08:15:55] <wjp|work> see my and Darke's previous comments
[08:16:02] <Cashman> ok
[08:16:18] <wjp|work> you got an anoncvs copy from in the middle of my commit
[08:16:48] <wjp|work> and anoncvs only updates every 24 hours or so :-(
[08:17:04] <Cashman> hmm ok I hardly understand - oh ok I see! didnt know it worked like that
[08:17:14] <wjp|work> it didn't... but it does now
[08:17:37] <wjp|work> http://sourceforge.net/docman/display_doc.php?docid=2352&group_id=1
[08:17:39] <Cashman> so even if I view the cvs on the net its a non developers ver - thats why ur committs in the irc logs arnt there
[08:17:43] <Cashman> thanks
[08:17:57] <DarkeSleep> http://sourceforge.net/docman/display_doc.php?docid=2352&group_id=1#cvs <- See this. They have a problem with their cvs server's performance and that's a temorary fix.
[08:18:15] <Cashman> oh ok I see
[08:19:29] <Cashman> I see thats why I saw somthing IIRC Darke "its the end of sourceforge, I knew it!"
[08:20:40] <DarkeSleep> Not really. It's just a joke on the fact that evertime sf.net seems to have problems, there's a few people who hop out of the woodwork and complain that sf.net is in dire straights and it's going to die Real Soon Now(tm).
[08:20:41] <-- Cashman has left IRC (Read error: 104 (Connection reset by peer))
[08:43:33] <Kirben> If you just want latest win32 binaries http://pentagram.sourceforge.net/snapshots/Pentagramwin32.exe
[08:43:57] <wjp|work> Kirben: you have devel cvs access right?
[08:44:06] <Kirben> yes
[08:46:51] <Kirben> Still too early to start regular snapshots ?
[08:47:02] <wjp|work> hm, don't know
[08:47:25] <Kirben> Latest cvs source code online at http://pentagram.sourceforge.net/snapshots/Pentagram.zip too
[08:47:28] <wjp|work> it would be ok by me, I guess, if you think it's interesting enough
[08:48:20] <DarkeSleep> I don't see a problem starting now or 'soon' anyway. We seem to have reached the break even point where lots of 'viewable' stuff is happening on a regular basis. *grin*
[08:49:08] <Kirben> ok just need to add links to web site.
[08:49:46] <wjp|work> DarkeSleep: ever heard of a 'istdiostream'?
[08:50:50] <wjp|work> I have some (apparently rather old) C++ code here (geomview-1.8.1) which seems to use it
[08:57:35] <DarkeSleep> Umm... *thinkthinkthink* I can't recall ever seeing such a beast.
[08:58:33] <DarkeSleep> It sounds like a handmade wrapper around stdout/stdin.
[09:17:59] <wjp|work> it doesn't look handmade
[09:18:25] <wjp|work> but the call to the constructor takes a FILE*, so it does look like a wrapper around stdio
[09:20:53] <DarkeSleep> Old gcc had a iostream that was effectively a wrapper around a FILE*. It's possibly from that.
[09:22:12] <DarkeSleep> What's the various timestamps on the code you're looking at?
[09:37:00] <wjp|work> feb 2001
[09:38:16] <DarkeSleep> That's just bizzare. Does it actually compile with gcc2.95.x?
[09:38:26] <wjp|work> no idea
[09:38:31] <wjp|work> don't have one of those around
[09:38:51] <wjp|work> oh wait, the solaris machine of the CS department has a gcc 2.95.x
[09:39:14] <DarkeSleep> emerge it, then use gcc-config to select the appropriate one? *grin*
[09:39:23] <wjp|work> no thanks
[09:39:27] <wjp|work> (besides, I'm not at home)
[09:39:50] <DarkeSleep> Awwww! Don't want the challenge? *grin*
[09:39:56] * wjp|work thinks
[09:39:58] <wjp|work> um, no ;-)
[09:40:15] <wjp|work> no istdiostream.h on that solaris machine
[09:40:56] <DarkeSleep> Done a search on the web to see if you can find something else that references it? It might have come from a compatability library from wayback.
[09:41:18] <wjp|work> ah, on an old linux machine from CS there _is_ a stdiostream.h
[09:41:53] <DarkeSleep> What gcc version? *grin*
[09:42:01] <wjp|work> RH7.3's
[09:42:27] <DarkeSleep> The evil 2.96?
[09:42:33] <wjp|work> yes
[09:45:35] * DarkeSleep doesn't think we can blame the presense of that header on redhat though. *grin*
[10:16:19] --> Cashman has joined #pentagram
[10:17:20] <wjp|work> ok... I think it's fairly easy to reproduce
[10:17:31] <wjp|work> there's apparently a 'stdio_filebuf' streambuf
[10:17:46] <wjp|work> and that should usable as underlying buffer for a stdiostream
[10:24:10] <wjp|work> ...although it doesn't have all the functions that geomview uses
[10:24:13] * wjp|work sighs
[10:24:33] <wjp|work> I'm starting to think this isn't worth the trouble :-)
[10:24:58] <DarkeSleep> What precicely is geomview, and why hasn't it been updated? *grin*
[10:31:30] <Cashman> 2003-06-14 - how many committs have been made ruffly after this date?
[10:31:58] <Cashman> I have no clue as to when I might recieve the valid update from the server
[10:33:27] <DarkeSleep> You can either download the source tarball that Kirben pointed to above, or wait 12-24 hours for the anon-cvs to update.
[10:34:13] <Cashman> ok thanks
[10:38:20] <Cashman> hmm so its just how cvs works! nothing to do with storing files in general I see
[10:38:36] <Cashman> the sourceforge tmp thing
[10:44:35] <Cashman> got the tarball
[10:49:57] <Cashman> yow I see there were lots of committs
[11:15:42] <-- Kirben has left IRC (Read error: 54 (Connection reset by peer))
[11:15:48] --> Kirben2 has joined #pentagram
[11:16:09] <Cashman> hmm items in gumps arnt doing anything
[11:16:51] <Cashman> ok wrong mouse buttom
[11:16:55] <Cashman> its ok
[11:27:36] <wjp|work> ok, Dominus is typoing on avatarInStasis too it seems :-)
[11:27:47] <wjp|work> that word appears to be quite hard :-)
[11:31:03] <wjp|work> DarkeSleep: www.geomview.org, someone here requested it to be installed
[11:40:48] <Cashman> ok I see when you demo to pythos the npc (mage) can walk (animate) sometimes, also pathfinds through the bookcase occasionally lol
[11:40:57] <Cashman> its looking good though!
[11:41:21] <wjp|work> pathfinding is a big hack currently
[11:41:25] <Cashman> oh I was wondering about the mirror reflection - I see there is a reflector shape
[11:41:41] <wjp|work> probably when you step on that it'll change the mirror to show your reflection
[11:41:47] <wjp|work> stepping on things isn't implemented yet, though :-)
[11:42:00] <Cashman> yeah I get it
[11:42:06] <Cashman> true true
[11:42:30] --- Kirben2 is now known as Kirben
[11:42:51] <Cashman> oh in u8 didnt text appear around (next to) other text if you clicked more than one thing
[11:42:56] <wjp|work> yes
[11:43:01] <Cashman> if it was possible? hmmm
[11:43:04] <wjp|work> above, actually
[11:43:20] <Cashman> so thats colourless domain right!?
[11:43:31] <wjp|work> hm, no, I meant two texts for the same item
[11:43:49] <wjp|work> not sure what happened when you get overlapping text for two items
[11:43:56] <wjp|work> you could check in the original? :-)
[11:43:58] <Cashman> ok
[11:44:16] <Cashman> I'll get that patch working again
[11:44:27] <Cashman> for win32
[11:44:41] <Cashman> yeah thats why I'm gonna get the patch to get it to work again
[11:46:44] <Cashman> will item dragging be implemented soon? I mean visible to the player
[11:46:51] <wjp|work> maybe
[11:47:00] <wjp|work> all the ingredients are there
[11:47:45] <Cashman> true
[11:48:38] <Cashman> anyway its nice and funny at the same time to see the progress stats change
[11:48:54] <Cashman> I mean not looking at the sourceforge ones - just see how progress is running generally
[11:51:18] <wjp|work> talking about stats... we have had pageview-spikes the last three saturdays
[11:51:26] * wjp|work wonders why
[11:51:37] <wjp|work> (usually about 50 hits/day. On saturdays about 120 hits/day)
[11:51:58] <Cashman> geez
[11:52:08] <Cashman> I know there are heaps of random view on sourceforge
[11:52:13] <Cashman> I like to do random views sometimes
[11:55:50] * Cashman goes to check text in u8 man I love the graphics in 640x480 its like a whole new game and more on screen at once
[11:56:00] <Cashman> I shudder when I run u8 again in 320x200 - wjp
[11:56:35] <Cashman> eeew suck horrible crash
[12:03:37] <Cashman> yeah wjp it put text as close as it could to the object (text around each other)
[12:03:57] <wjp|work> no overlaps?
[12:04:13] <Cashman> so far I see none!
[12:04:28] --> Colourless has joined #Pentagram
[12:04:31] --- ChanServ gives channel operator status to Colourless
[12:04:31] <wjp|work> does text overlap gumps?
[12:04:47] <wjp|work> hi
[12:05:00] <Colourless> hi
[12:05:16] <Colourless> wjp: lsh.... is that correct?
[12:05:22] <wjp|work> no idea; is it?
[12:05:45] <wjp|work> no, it looks rather wrong :-)
[12:06:27] <Colourless> :-)
[12:06:31] <wjp|work> committed
[12:07:34] <Cashman> no text doesnt overlap on gumps
[12:07:55] <Cashman> but sometimes the text is drawn behind the gump e.g. the basket where devon's camp is
[12:08:10] <Cashman> icky
[12:08:47] * wjp|work wonders which (if any) known bugs will be fixed by that
[12:09:29] <Colourless> properly aligning multiple bark gumps will not happen soon :-)
[12:09:49] <wjp|work> aww :-)
[12:09:52] <Colourless> it's just 'not important' :-)
[12:10:30] <wjp|work> I wonder if you can get them to overlap when you move items around while they have open BarkGumps
[12:11:05] * Cashman is just talking crazy because he feels like it, some nutter might want to try and make the lost vaule now
[12:11:46] <Colourless> um, the 0-9 A-E key from readme need removing.
[12:11:49] <Cashman> not many know of pentagram so I guess no nutters yet
[12:11:52] <Colourless> those keys don't do that anymore :-)
[12:12:23] * Cashman notes hes running ultima8 right now but its slow in windows
[12:12:40] <wjp|work> Colourless: good point :-)
[12:12:41] <wjp|work> committed
[12:13:38] <Colourless> next thing on the todo list: limit bark gumps to only 5 lines of text
[12:13:45] * wjp|work nods
[12:13:49] * Cashman was wondering if oneday if sfx gets a look at again, if its worth decoding all the sfx using the origional .exe in one go and putting them into a sound.flx like TGWDS
[12:14:11] <wjp|work> huh?
[12:14:22] <wjp|work> parse error :-)
[12:15:25] * Colourless points at dcmpsound
[12:18:09] <Colourless> sfx is a non issue
[12:18:33] <Cashman> yeah so I will stop talking about it
[12:18:39] <wjp|work> apart from figuring out what the bazillion sfx intrinsics do :-)
[12:18:55] <Cashman> hehe
[12:20:48] <Colourless> found out this a little earlier. U8's time is 4x accelerated. So 1 game hour is 15 real minutes
[12:23:06] <Colourless> or 27000 ticks per game hour
[12:24:35] <Colourless> wjp, spawn/spawn inline shouldn't be setting cede. The suspend instruction will implicitly set cede IF it's wanted
[12:24:53] <wjp|work> brb
[12:25:21] <wjp|work> that'll give problems with string/list params
[12:25:34] <Colourless> because with short spawned processes that don't spawn anything else, will execute, terminate, and their pid will become invalid BEFORE the implies opcode in the caller is done
[12:25:52] <Colourless> so the caller can not get the return
[12:27:22] <Colourless> now, there might be a posssible work around
[12:27:40] <Colourless> since i forgot about the string/list stuff
[12:28:15] <Colourless> terminated processes should wait a frame before having their pid cleared
[12:28:31] <Colourless> possibly just changing this in run
[12:28:33] <Colourless> if (!p->terminated)
[12:28:33] <Colourless> if (p->run(framenum)) dirty = true;
[12:28:33] <Colourless> if (p->terminated) {
[12:28:34] <Colourless> to
[12:28:43] <Colourless> if (!p->terminated)
[12:28:43] <Colourless> if (p->run(framenum)) dirty = true;
[12:28:43] <Colourless> else {
[12:29:57] <Colourless> uh, not quite :-)
[12:30:00] <Colourless> nested if's
[12:33:57] <Colourless> hmm, no, that's not working
[12:39:47] <Colourless> i don't know the solution
[12:43:20] <Colourless> my previous 'execute spawned processes' imediately actually did work here
[12:43:56] <Colourless> s/imediately/imediately hack/
[12:44:28] <Colourless> cause implies would always be called before the kernel had an opportunity to delete the process
[12:44:42] <wjp|work> I think that's a bit too much of a hack
[12:46:58] <Colourless> perhaps, but until it gets fixed, the Schedule events for the npcs just wont work
[12:48:09] <wjp|work> maybe pass strings/lists by value instead of by ref.
[12:48:15] <wjp|work> and then don't cede
[12:48:58] <Colourless> thats not really possible
[12:49:04] <Colourless> we don't know what is or isn't a string
[12:50:14] <wjp|work> hm, good point
[12:51:54] <Colourless> btw for the Schedules all of them spawn 057C:143A, which executes, terminates and is deleted by the kernel before implies, suspend and push process result. This happens
[12:51:56] <Colourless> sp = 1A; 01C1: 54 impliesNon-existant process PID (960) in implies
[13:00:45] <wjp|work> we could just delay deleting a terminated process
[13:00:54] <wjp|work> although that's somewhat hackish too
[13:03:18] <Colourless> i'm trying that now. ret from process will not terminate till the next frame
[13:18:32] <wjp|work> any luck?
[13:23:32] <Colourless> yeah that works
[13:24:22] <Colourless> now though i want a nice way of implementing it :-)
[13:25:59] <Colourless> should it be something that should be able to be applied to all process types, or only UCProcesses?
[13:26:36] <Colourless> i.e. should it be a feature of the kernel and process, or just ucprocess
[13:29:07] <Cashman> http://www.shiftedphase.com/~dino/ultima/uinfo/u8/u8fun.php
[13:29:18] <Colourless> have a flag terminate_deferred or somehing
[13:29:34] <Cashman> the part "a close looked at the execution scene" - just read later
[13:29:47] * Cashman Cashhiding
[13:29:54] --- Cashman is now known as CashZzz
[13:35:45] <wjp|work> I'd say all processes
[13:36:18] <Colourless> yeah, i'm adding in the functionality now
[13:39:39] <CashZzz> http://www.shiftedphase.com/~dino/ultima/uinfo/u8/util/u8save.zip -hehe
[13:39:39] <-- CashZzz has left IRC ()
[14:08:35] <Colourless> wjp, any issues if i add a WorldPoint class for use with the intrinsics?
[14:08:44] <wjp|work> not at all :-)
[14:09:14] <Colourless> just a tiny thing that has a buffer of 5 bytes, and 3 inlined functions to get x y and z :-)
[14:11:44] <Colourless> also added a few useful macros
[14:11:51] <Colourless> such as ARG_OBJECT_FROM_ID :-)
[14:11:54] <wjp|work> :-)
[14:12:24] <Colourless> i'm thinking of renaming the other ARG_OBJECT and similar macros to ARG_OBJECT_FROM_PTR
[14:12:38] <wjp|work> hm... how can a shape with translucency be occluding?
[14:12:40] <wjp|work> (shape 40)
[14:13:02] <wjp|work> ok by me, if you feel like renaming all of the existing ones
[14:15:02] <Colourless> that doesn't make much sense logically if it is
[14:15:49] <Colourless> but it claims it does :-)
[14:15:52] <wjp|work> 40, 0: Info: Fixed Solid Land Occl Trans Generic
[14:16:22] <Colourless> yeah. i'm thinking mistake
[14:16:48] <wjp|work> occl = occlude && !invisible && !translucent? :-)
[14:23:58] <Colourless> :-)
[14:24:02] <Colourless> yes
[14:35:09] <-- Kirben has left IRC (Read error: 54 (Connection reset by peer))
[14:44:34] <Colourless> ok, committed my few changes. nothing to major, execpt NPC schedules are now executed on map changes.
[14:44:57] <Colourless> Remember, do not wake up the tempest :-)
[14:45:14] <wjp|work> that was fatal, wasn't it? :-)
[14:45:25] <Colourless> indeed it was :-)
[14:45:33] <Colourless> of course in pentagram you are just in stasis :-)
[14:45:42] <Colourless> since we don't support death yet :-)
[14:47:08] <Colourless> now if you hold up and right in central tenebrae after the execution, you'll go straight through mordeas bedroom, and she'll zap you :-)
[14:47:44] <wjp|work> yay :-)
[14:47:50] <wjp|work> how... convenient :-)
[14:48:29] <wjp|work> what does cachein do to the start sequence?
[14:48:42] <Colourless> avatar gets up right away
[14:48:50] <wjp|work> without first being run?
[14:48:54] <Colourless> yes
[14:49:12] <Colourless> first will still run if you press 'f' and the avatar gets up... again :-)
[14:51:37] <Colourless> cachein itself doesn't do much.
[14:51:41] <Colourless> only 2 classes have it
[14:51:45] <Colourless> Avatar and Music :-)
[14:52:04] <Colourless> for Music, it plays music, duh, and for avatar it disabled air walk
[14:52:28] <Colourless> and possibly some other things as awll
[14:53:33] <Colourless> hmm,
[14:53:48] <Colourless> avatar will not get up if the 'getup' global is set to 1
[14:54:17] <Colourless> which gets set by cachein
[14:54:29] <Colourless> so why does he get up a second time...
[14:56:30] <Colourless> hmm, because the avatar must get up for some other reason
[14:57:10] <Colourless> forcing global 0x10 to 1 in UCMachine::UCMachine() stops av from getting up due to cachein
[14:57:32] <wjp|work> hm, are globals stored in initgame.dat?
[14:58:35] <Colourless> flag.dat?
[14:58:41] <wjp|work> how big is it?
[14:58:51] <Colourless> 1 k
[14:58:56] <wjp|work> and does it have a 1 at position 0x10? :-)
[15:00:08] <Colourless> it's not in the initial savegame
[15:01:34] <wjp|work> where did you hide WorldPoint.h? :-)
[15:04:11] <Colourless> added :-)
[15:04:19] <Colourless> it's in world btw :-)
[15:06:51] <Colourless> forcing a save right at the start of the game produces a flag.dat that is completely empty
[15:07:21] <wjp|work> time to go home
[15:07:22] <wjp|work> bbl
[15:07:28] <Colourless> cya\
[15:07:30] <-- wjp|work has left IRC ("bbl")
[15:32:12] --> wjp has joined #pentagram
[15:32:12] --- ChanServ gives channel operator status to wjp
[15:33:39] <Colourless> wb
[15:34:07] <wjp> thx
[15:40:20] <Colourless> each period of time (bloodwatch,firstebb) lasts 4 game hours, which actually lasts 1 real hour.
[15:40:44] <Colourless> no wonder time seemed to move slowly in pagan :-)
[15:41:13] <wjp> indeed :-)
[15:41:20] <wjp> thank god for bedrolls :-)
[15:41:30] <Colourless> which don't work yet :-)
[15:41:43] <wjp> they don't advance time, no :-)
[15:41:50] <wjp> (since there is no time, but who cares ;-) )
[15:41:56] <Colourless> there is now :-)
[15:42:16] <Colourless> my commit added I_getTimeInGameHours
[15:42:18] <wjp> ah, nice :-)
[15:43:37] <Colourless> time advances really slowly though :-)
[15:58:20] <wjp> item placement idea: trace object below cursor, and place dragged item next to the traced item in the 'direction' of the mouse
[15:58:52] <wjp> I hope that will give intuitive results
[15:59:08] <Colourless> hmm, it depends on where the mouse hit
[15:59:28] <Colourless> if the mouse hit the top of the object's bounding box, you would think to place on top
[15:59:38] <wjp> yes, which is what should happen then
[15:59:38] <Colourless> of it it the end, then you'd think next to it
[15:59:54] <wjp> if you drag it onto the side of an object, it will move there and then fall
[16:02:52] <Colourless> ok original worked like this
[16:03:10] <Colourless> if you were within 'reaching' distance, you could only place an object on top of something else
[16:03:34] <Colourless> dragging to the edge of an object wasn't allowed (it would bring up the red x)
[16:03:56] <Colourless> if you were throwing an object, you could throw it at the side of another object. it would then bounce off
[16:04:38] <wjp> how accurately can you throw items?
[16:05:03] <Colourless> it was pretty good, but they bounced :-)
[16:05:34] <wjp> so if you aim for the ground, it lands where you aimed but then moves on?
[16:06:05] <Colourless> depends. it seems like it's a bit random
[16:06:14] <Colourless> sometimes it doesn't hit the right spot
[16:06:33] <wjp> some randomness makes sense
[16:06:39] <wjp> I wonder if it depends on your dex :-)
[16:06:50] <Colourless> the bounce was also random
[16:07:09] <Colourless> i just threw a basket and it bounced back towards me
[16:07:23] <wjp> eh?
[16:07:34] <wjp> after it hit a flat, horizontal surface?
[16:07:39] <Colourless> hit the ground and bounced back a short distant
[16:07:55] <Colourless> not far though, only a few pixels
[16:08:32] <Colourless> i'm guessing the boucing always occurs in the direction of the throw, but it can have a negetive multiplier
[16:09:35] <wjp> or maybe there's a random interval around the 'target' where it can land?
[16:09:53] <wjp> (and that interval can semi-accidentally be longer than the distance it should bounce)
[16:10:46] <Colourless> well, the throw function or whatever calculates a parabola from avatar to the dest
[16:11:04] <Colourless> probably using hurl
[16:12:31] <wjp> the game probably continues to move as soon as the item starts moving, right?
[16:12:39] <Colourless> yes
[16:12:44] <wjp> good :-)
[16:13:01] <Colourless> i found out the hurl is a process btw :-)
[16:13:19] * wjp nods; I read the commit msg
[16:14:51] <Colourless> hurl has 4 args...
[16:15:14] <Colourless> usually range from about -16 to 16
[16:15:47] <wjp> coefficients for a polynomial? :-)
[16:15:48] <Colourless> though not always
[16:15:55] <Colourless> i think so :-)
[16:16:21] <Colourless> but 3 of them might for a velocity vector
[16:17:12] * wjp nods
[16:57:01] <wjp> dinner's ready; I'll bbl
[17:24:23] <-- oreilly has left IRC ()
[17:28:44] <wjp> back
[17:28:54] <Colourless> wb
[17:32:47] <Colourless> there is some comment about temporary npcs somewhere and 'flags'
[17:32:55] <Colourless> i think they are supposed to have the FLG_MONSTER_NPC set
[17:33:26] <wjp> how am I supposed to read that flgname? "an npc of type monster"?
[17:33:48] <wjp> "monster or npc"?
[17:34:19] <Colourless> npc spawned by monster egg
[17:35:24] <wjp> ok, added that
[17:49:44] <wjp> gah, I'm tired
[17:49:55] <wjp> sleeping 4 1/2 hours is _not_ good :-)
[17:49:59] <Colourless> so am i :-)
[17:50:13] <wjp> not even 8pm here, though :-)
[17:50:43] <Colourless> i myself are going in a few minutes :-)
[17:51:09] <wjp> hm, the time looks suspiciously like "real" seconds
[17:51:41] <Colourless> it is real seconds
[17:51:47] <Colourless> 15 minutes per hour
[17:51:59] <wjp> ok
[17:51:59] <Colourless> that is 15 real minutes per game hour
[17:52:15] <wjp> 4 hours per period, 6 periods per day, right?
[17:52:16] <Colourless> the usecode execpts that (in all of the 1 functions that calls for minutes)
[17:52:22] <Colourless> yes
[17:53:45] <Colourless> i'm off
[17:53:48] <Colourless> cya
[17:53:50] <wjp> night
[17:53:51] <-- Colourless has left IRC ("casts invisibility")
[19:03:29] --> SB-X has joined #pentagram
[20:18:45] <-- DarkeSleep has left IRC (kornbluth.freenode.net irc.freenode.net)
[20:19:16] --> DarkeSleep has joined #pentagram
[20:19:17] --- ChanServ removes channel operator status from DarkeSleep
[22:05:18] <wjp> middle mouse 'hurl's an item now :-)
[22:05:21] <wjp> endless fun :)
[22:08:25] <DarkeSleep> Ooooh... so we can throw guards around the place?
[22:08:30] <wjp> um, yes :_)
[22:09:01] <DarkeSleep> Yay!
[22:09:29] <DarkeSleep> "Take that! *click!*" "Aieeeeeee!"
[22:10:19] * wjp throws Devon up a tree
[22:10:27] * wjp throws the tree
[22:11:16] <wjp> you can use it to climb onto roofs too
[22:12:54] <wjp> hm, darts 'kind of' work
[22:13:23] <wjp> sometimes
[22:13:47] * wjp double-checks he committed everything
[22:13:54] <wjp> ok, looking good :-)
[22:13:59] <wjp> time to go; bye
[22:14:06] <-- wjp has left IRC ("Zzzz...")